MATERIALS AND METHODS

Sample collection and isolation of lactic acid bacteria: Samples of vaginal swabs from healthy and fecund females of reproductive age group were collected were transferred immediately to sterilized saline (0.85% NaCl). The sample was inoculated into sterile MRS broth (Demann Regossa Sharpe number 1164 with 1 ml L^-1. tween-80 pH was adjusted 6.5) for propagating vaginal LAB microflora. After incubation of 18-24 h at 37°C. Samples were subcultured three times before proceeding with bacterial isolation and activity assays. Lactic Acid Bacteria (LAB) strains were isolated on MRS agar plates using pure culture techniques.

Bacterial strains and culture media used to study antimicrobial spectrum: The inhibitory spectra of bacteriocin producing human vaginal LAB isolates was evaluated against important pathogens especially associated with Bacterial Vaginosis, using spot-on-lawn (Toba et al., 1991) and well-diffusion methods (Pucci et al., 1988). Growth requirements of indicator microorganisms Associated with BV are specified in Table 1. Indicator strains were revived and maintained in growth media as prescribed by culture banks. Indicator cultures have been procured from Microbial Type Culture Collection, Chandigarh, Punjab, India Bacteroides fragilis MTCC 1045, Bacteroides ovatus MTCC 3298, Bacteroides vulgatus MTCC 1350, Candida albicans MTCC 183, Neisseria mucosa MTCC 1772, Staphylococcus aureus MTCC 737, Streptococcus faecalis MTCC 459, Streptococcus thermophilus MTCC 1928, National Collection of Industrial Microorganisms, Pune, India Proteus mirabilis NCIM 2387, Streptococcus agalactiae NCIM 2401 and from American Type Culture Collection G. vaginalis ATCC 10418, Micrococcus flavus ATCC 10240, Neisseria gonorrhoeae ATCC 19424 Candida albicans ATCC 10231 Staphylococcus albus ATCC 11631 All other cultures including Staphylococcus aureus NCTC 7447, Staphylococcus pyogenes NCTC 1086914 were kindly provided by Orbit Biotech, Mohali, Punjab, India. Gardnerella vaginalis ATCC 14018 was revived and maintained in the Casman’s medium containing Gardnerella vaginalis active supplement (constituting gentamicin sulphate, nalidixic acid and amphotericin B) and 5% w/v defibrinated human blood.

Bacteriocin activity assays: Isolated strains were subcultured thrice in MRS medium (pH 6.5) at 37°C for 24 h before proceeding with bacteriocin activity assays. Out of 10 of 1 mL aliquot was centrifuged at 10,000 rpm for 10 min and Cell Free Supernatant (CFS) was collected in sterile micro-centrifuge tube. CFS was heat treated in boiling water bath for 20 min and allowed to cool at room temperature. Bacteriocin activity was assayed using spot-on-lawn and agar well diffusion methods.

Determination of viability loss
Turbidity method: Cell suspensions of Gardnerella vaginalis of organisms were grown as broth cultures (Casman broth+selective media) along with increasing concentrations of the bacteriocin to be screened and turbidity at 600 nm was measured after 24-48 h. The degree of turbidity in the broth culture is directly related to the number of microorganism present, either viable or dead cells. Thus, the increasing the turbidity of the broth medium indicates increase of the microbial cell mass.

Colony count method: The cell suspensions of Gardnerella vaginalis were used to prepare an experimental set of samples as follows: Untreated control, test samples treated with 0-500 μg mL^-1 pure bacteriocin. After 24 h incubation at 37°C, experimental set was examined to enumerate CFU mL^-1. Enumeration was done by pour plating original suspension or serially diluting them with sterile saline and incubating them on MRS agar plates at 37°C for 24 h. Results of bacteriocin treated G. vaginalis suspensions were compared with untreated control. A reduction in cell counts mL^-1 following bacteriocin treatment was regarded as cell viability loss (Kalchayanand et al., 2004).

SEM analysis of bacteriocin-treated G. vaginalis to indicate its mode of action: The control and treated cells of the bacterial vaginosis pathogen Gardnerella vaginalis were examined by scanning electron microscope model-hitachi VP-SEM S-3400N, Europe, to visualize morphological change that occurred in the cells following the exposure to bacteriocin produced by human vaginal isolate HV6b. Gardnerella vaginalis cell suspension was prepared in distilled water and it was treated with pure bacteriocin at a concentration of 0.5 mg mL^-1. Suspension was kept for 90 min at 37°C in a hydrated chamber and then centrifuged at 10,000 rpm for 1 min at 4°C. Cell pellet was dissolved in minimum quantity of ethanol and a drop of it was transferred to poly-L-lysine-treated silicon wafer chips, that were kept for 30 min in a hydrated chamber for the cells to adhere. Chips were viewed at 10 kV accelerating voltage in a field emission scanning electron microscope and images of the bacteriocin treated cells for topography contrast were collected at different magnifications.

Investigation of antibiotic susceptibility: Antibiotic susceptibility Lactobacillus fermentum and G. vaginalis was studied against different concentrations (2-8 μg mL^-1) of some commonly prescribed antibiotics to BV patients including ampicillin, amoxicillin, amoxicillin and clavulanic acid, azithromycin, ciprofloxacin, co-trimoxazole, erythromycin, gentamicin, metronidazole, nalidixic acid, ofloxacin, penicillin, rifampicin, tetracycline, tinidazole and vancomycin by spot-on-lawn method using.

Statistical analysis of data: QI Macros ANOVA statistical tools like ANOVA one factor was used to determine significance of the results obtained in duplicate experimental sets. Level of significance was adjusted at 0.05 and results were said to be significant if their p-value <0.05 and F_crit<F-value according to one way ANOVA.

RESULTS AND DISCUSSION

Isolation of bacteriocin producing LAB strains: One of the bacteriocin producing strain was selected from more than 100 bacterial strains that were isolated from vaginal swabs after screening on the basis of bacteriocin producing activity. Bacteriocin production was observed in only eight isolates designated as HV6, HV6b, HV75, HV76, HV54A, HV59A, HV59C, HV59D and HV69. Their antibacterial activity was tested against indicator strains implicated in BV (Table 1). Isolate HV6b possessed antimicrobial spectrum that was highest among the tested indicators especially G. vaginalis ATCC14018 and therefore, was selected for further detailed study.

Antimicrobial spectrum of the isolated LAB strains against BV associated organisms Antimicrobial activities of isolated LAB strains were investigated against a panel of microorganisms associated with bacterial vaginosis and highly significant results were obtained in the study (Table 2; Fig. 1, 2). Growth inhibition of Urinary Tract Infections (UTI) pathogen N. mucosa by bacteriocins of vaginal LAB isolates makes them potential ingredient of anti-neisserial skin/mucosal formulations to eradicate such opportunistic UTI pathogens (Reid, 2001). G. vaginalis causing bacterial vaginosis in humans was strongly inhibited by human vaginal isolates 6, 6b, 76 and 59A. Several investigators have isolated and partially purified bacteriocin from different species of lactobacilli. Most of them were with nonhuman strains, predominantly isolated from food (Karaoglu et al., 2003). Bacteriocin produced by LAB isolate HV6b has a great potential to control G. vaginalis associated BV in humans. The use of human lactobacilli as probiotics assigned to restore and maintain a healthy urogenital tract represents a promising alternative to conventional chemotherapy (Kaur et al., 2010; Kumar et al., 2011). All the tested strains of Bacteroides were inhibited by isolate HV6b.

Table 1:	Growth media and conditions of indicator microorganisms
F. anaerobe*: Facultative anaerobe, NB: Nutrient Broth, RCB: Reinforced Clostridial Broth, MRS: de Man’s Ragosa Sharpe Medium, CB: Casman’s Broth, BHI: Brain Heart Infusion Broth, CMM: Cooked meat media, YEPD: Yeast extract potato dextrose, BV: Bacterial vaginosis

Table 2:	Antimicrobial spectrum of human vaginal LAB isolates against BV associated pathogens
Each data is an average of two samples, p-value <0.05, Fcrit (1.7184)<F-value (17.966), *: Negative reaction

Fig. 1:

Spot-on-lawn bacteriocin activity assay

Fig. 2(a-b):

Well diffusion bacteriocin activity assay against indicator strains (a) Streptococcus faecalis MTCC 459 and (b)Gardnerella vaginalis ATCC 14018 showing prominent inhibition zones

B. fragilis is reported to cause obstetrics and gynecologic infections. B. ovatus infections can develop in all body sites, including the CNS, head, neck, chest, abdomen, pelvis, skin and the soft tissues (Sweet, 1985). B. vulgatus is commonly found in patients with ulcerative colitis. Inadequate therapy against these anaerobic bacteria may lead to clinical failure, thus probiotic therapy consisting of bacteriocin producing Lactobacilli is being suggested here (Bamba et al., 1995).

C. albicans, Mobiluncus, M. hominis, Peptostreptococcus spp. and P. bivia are some other BV associated pathogens of urinary tract. Inhibition of C. albicans was also studied that indicated moderate to high inhibition by bacteriocins of LAB origin especially isolate HV6b. An acidic pH of vagina alone is not sufficient to inhibit vaginal pathogens and to prevent bacterial vaginosis. Thus, bacteriocin based therapeutics are urgently desired to cure such diseases and to overcome problems associated with antibiotic therapy such as recurrence of vaginal infections. There are many evidence that indicate potential of Generally Recognized As Safe (GRAS) lactic acid bacteria in maintaining and restoring gut homeostatic (Thirabunyanon, 2011).

Table 3:	Percentage growth of bacteria with bacteriocin by turbidity method

Table 4:	Changes in colony counts of G. vaginalis following treatment with bacteriocin produced by HV6b isolate
Each data is an average of two samples, p-value <0.05, Fcrit (4.3877) <F-value (65.543)

Use of live probiotic bacteria may have prophylactic applications but use of purified bacteriocins appears to be more attractive for eradicating an established infection (Lohans and Vederas, 2012). Ideally, anti-BV or anti-diarrheal therapeutics should specifically target disease causing microorganism and should have least interference with health promoting commensally microflora. In fact, the spectrum of activity of L. fermentum HV6b may be extremely well suited for targeting specific pathogens in vivo. In contrast to it, antibiotic prescribed frequently to cure gut and vaginal infections, strongly inhibit most of these beneficial microorganism at much lower concentrations (Le Blay et al., 2007).

Viability loss in bacteriocin treated G. vaginalis samples
Turbidity method: Cell suspensions of G. vaginalis of organisms grown as broth cultures (Casman broth+selective media) with increasing concentrations of the bacteriocin showed a sharp decrease in growth as at 330 μg mL^-1 of bacteriocin 50% of turbidity was decreased and there was no growth of G. vaginalis with 500 μg mL^-1 of bacteriocin turbidity at 600 nm is measured after 24-48 h. At the concentration of at 500 μg mL^-1 no turbidity was found (Table 3).

Colony count method: The cell suspensions of G. vaginalis in MRS broth were treated with (0-500 μg mL^-1) different concentrations of pure bacteriocin and incubated at 37°C at 24 h. Gardnerella counts were determined and compared with respective control. Colony counts (log 10 CFU mL^-1) in the control samples were highest while reductions were observed in bacteriocin treated samples in a concentration dependent manner indicating the loss of viability. The count was reduced by 0.11 log cycles in G. vaginalis treated suspension with 100 μg mL^-1 bacteriocin but by over 1.29 log cycles in suspensions of G. vaginalis treated with 400 μg mL^-1 bacteriocin. There was a complete loss in viability and no CFU was detected in samples treated with 500 μg mL^-1 bacteriocin. Viability loss of pathogenic bacteria exposed to various bacteriocins has been reported by other researchers (Ray et al., 2001). It has been shown that bacteriocin produced HV6b isolate is highly potent anti G. vaginalis agent with numerous therapeutic applications. If we colony forming units with control there was not a single colony at 500 μg mL^-1 (Table 4).

Fig. 3(a-c):

Scanning electron micrographs of bacteriocin-treated G. vaginalis (a) Whole cell showing pore formation and (b) and (c) enlarged portions of the bacteriocin disrupted cell wall

SEM analysis of bacteriocin-treated G. vaginalis to indicate its mode of action: The SEM generated photo-micrographs of the control and treated cells of the pathogen G. vaginalis are presented in Fig. 3a-c; clear changes in cell morphology can be seen. The bacteriocin treated cells appeared collapsed that could be associated with the damage in the cell wall and cell membrane followed by lyses. Disruption of the cell wall and cell membrane is due to formation of a large number of pores (of small to moderate size) causing extensive damage to the cells and thus, inducing viability loss. These findings agree with the results of viability loss in bacteriocin treated G. vaginalis cells. The results clearly show that size of pores was increasing with increasing concentration of Bacteriocin.

Antibiotic susceptibility of vaginal Lactobacillus fermentum and G. vaginalis: Vaginal isolates displayed variations in their sensitivity to commonly prescribed antibiotics for treating BV. The safety investigation of the bacterial vaginal isolates revealed that they were sensitive to antibiotics viz. ampicillin, amoxicillin and clavulanic acid and penicillin but express a natural resistance phenotype to amoxicillin, azithromycin, erythromycin, gentamicin, metronidazole, nalidixic acid and tinidazole. Maximum inhibition zone was observed in case of isolate HV6 against ampicillin (20 mm) followed by amoxicillin and clavulanic acid (19 mm). Antibiotic susceptibility of vaginal isolates against various antibiotics is given in Table 5. HV6b, HV54A and HV75 exhibited resistance to ampicillin. Strain HV69 showed its resistance to amoxicillin and clavulanic acid and its sensitivity to amoxicillin, azithromycin and metronidazole. On the other hand, G. vaginalis exhibited sensitivity to amoxicillin and clavulanic acid, ciprofloxacin, ofloxacin, penicillin, tetracycline, tinidazole and rifampicin at more than 2 μg mL^-1 concentration, to cefixime, metronidazole and metronidazole H at a concentration more than 4 μg mL^-1 and to co-trimoxazole and erythromycin at higher concentration (above 8 μg mL^-1). Resistances of HV6b isolate to most of these antibiotics except for penicillin and amoxicillin and clavulanic acid indicated its suitability as a component of triple therapy for bacterial vaginosis (Table 5) even we have compared the zone of bacteriocin HV6b against Gardnerella vaginalis along with the table of antibiotics which clearly indicates better and safe therapy with bacteriocin.

Table 5:	Antibiotic susceptibility of G. vaginalis ATCC14018 and L. fermentum HV6b where R and ND refers to resistant and not defined
R: Resistance, nd: Not defined

Fig. 4(a-b):

Comparative GC chromatogram showing lactic acid production in (a) Cell free supernatant, (b) reference lactic acid

In a review on antibiotic sensitivity pattern of Gardnerella vaginalis which was done with cefuroxime and ceftazidime, ceftriaxone, cloxacillin, erythromycin, chloramphenicol and metronidazole. Metronidazole was used as the first-line drug of choice for the treatment of G. vaginalis (Adinma et al., 1997). Nagaraja (2008) studied resistance of Gardnerella vaginalis with and recurrence of bacterial vaginosis after the treatment with metronidazole.

Characterization of Lactobacillus fermentum: In GC analysis, the polyethylene glycol (PEG) column was used. Lactic acid production was confirmed by gas chromatography by comparing it with a standard lactic acid (Fig. 4).

CONCLUSIONS

A highly potent anti-Gardnerella vaginalis bacteriocin producing isolate from human vaginal ecosystem was characterized based on its antimicrobial spectrum. Bacteriocin production trait of L. fermentum HV6b isolate was studied against important human pathogens causing bacterial vaginosis. This bacteriocin is a novel protein. A concentration dependent viability loss, tributary method and disruption of bacterial cells were observed in G. vaginalis culture exposed to bacteriocin produced by HV6B isolate that was further confirmed by SEM analysis. Antibiotic susceptibility of the isolated Lactobacillus fermentum and G. vaginalis was tested against many commonly prescribed antibiotics. Based on the results obtained in this study, human vaginal LAB isolates are strongly recommended for formulating anti-BV therapeutics/ointments/vaginal creams aimed at prophylaxes and treatment of BV and other sexually transmitted diseases in combination with antibiotic therapy that could check recurrence of the disease after termination of antibiotic treatment. Growing scientific evidences have proven efficacy of probiotics. Isolate HV6b showed inhibition of most of the tested pathogens. But it did not interfere with most of the Gram-positive probiotic lactic acid bacteria tested in the study. Therefore, there is a least possibility of its interference with normal human microflora, in contrast to frequently prescribed antibiotics. Preliminary experiments have established bacteriocin produced by L. fermentum sp. HV6b as an effective antimicrobial agent, with very impressive market value to formulate personal care products.

ACKNOWLEDGMENTS

Special thanks to Prof. Raj Bahadur Director Principal Govt. Medical College and Hospital (GMCH), Sector 32, Chandigarh, India, Prof. Anju Huria, Head of Department Obstt. and Gynae, GMCH, Chandigarh and Nurse Uma and Dr. Darshan Jot, Gynecologist, Phase 10, Mohali for their support in collecting samples under mutual consent. I also want acknowledge, Dr. Inderpreet Kaur, Scientist, Central Scientific Instrument Organization (CSIO), Chandigarh, India for extending facility of scanning electron microscopy. I also thankful to Jagvir Singh, Scientist (CSIO).